The Lysosomal Transmembrane Protein 9B Regulates the Activity of Inflammatory Signaling Pathways*

The intracellular signaling pathway by which tumor necrosis factor (TNF) induces its pleiotropic actions is well characterized and includes unique components as well as modules shared with other signaling pathways. In addition to the currently known key effectors, further molecules may however modulate the biological response to TNF. In our attempt to characterize novel regulators of the TNF signaling cascade, we have identified transmembrane protein 9B (TMEM9B, c11orf15) as an important component of TNF signaling and a module shared with the interleukin 1β (IL-1β) and Toll-like receptor (TLR) pathways. TMEM9B is a glycosylated protein localized in membranes of the lysosome and partially in early endosomes. The expression of TMEM9B is required for the production of proinflammatory cytokines induced by TNF, IL-1β, and TLR ligands but not for apoptotic cell death triggered by TNF or Fas ligand. TMEM9B is essential in TNF activation of both the NF-κB and MAPK pathways. It acts downstream of RIP1 and upstream of the MAPK and IκB kinases at the level of the TAK1 complex. These findings indicate that TMEM9B is a key component of inflammatory signaling pathways and suggest that endosomal or lysosomal compartments regulate these pathways.

The intracellular signaling pathway by which tumor necrosis factor (TNF) induces its pleiotropic actions is well characterized and includes unique components as well as modules shared with other signaling pathways. In addition to the currently known key effectors, further molecules may however modulate the biological response to TNF. In our attempt to characterize novel regulators of the TNF signaling cascade, we have identified transmembrane protein 9B (TMEM9B, c11orf15) as an important component of TNF signaling and a module shared with the interleukin 1␤ (IL-1␤) and Toll-like receptor (TLR) pathways. TMEM9B is a glycosylated protein localized in membranes of the lysosome and partially in early endosomes. The expression of TMEM9B is required for the production of proinflammatory cytokines induced by TNF, IL-1␤, and TLR ligands but not for apoptotic cell death triggered by TNF or Fas ligand. TMEM9B is essential in TNF activation of both the NF-B and MAPK pathways. It acts downstream of RIP1 and upstream of the MAPK and IB kinases at the level of the TAK1 complex. These findings indicate that TMEM9B is a key component of inflammatory signaling pathways and suggest that endosomal or lysosomal compartments regulate these pathways.
Tumor necrosis factor (TNF) 2 is a pleiotropic mediator of a wide range of cellular responses to infection, such as cytokine and chemokine production, cell migration, cell death, and cell differentiation and maturation (1). TNF plays a pivotal role in several autoimmune disorders such as rheumatoid arthritis; this is underscored by the clinical success of neutralizing TNF with antibodies or soluble receptors (2). A better understanding of intracellular TNF signaling is therefore of high clinical relevance.
The two TNF receptors, TNFR1 (p55, TNFRSF1A) and TNFR2 (p75, TNFRSF1B), show high homology in their extra-cellular domains but less in their intracellular domains. Although soluble TNF binds TNFR1 with higher affinity than TNFR2 and therefore acts primarily via TNFR1, membranebound TNF activates equally TNFR1 and TNFR2 (3). In most tissues TNF signaling is mediated by TNFR1, whereas TNFR2 is restricted to fewer specific tissues, mostly of an immunological nature (3). Upon ligand binding, TNFR1 trimerizes and recruits TNF receptor-associated death domain protein (TRADD), receptor-interacting protein 1 (RIP1), and TNF receptor-associated factor 2 (TRAF2). This first complex acts as a platform at the plasma membrane to activate the NF-B and MAPK signaling cascades, promoting cell survival and the expression of inflammatory cytokines. In a second step, TNFR1 is internalized into endocytic vesicles together with TRADD and RIP1 and recruits the proapoptotic molecules Fas-associated death domain (FADD) and caspase-8. This complex will initiate the apoptotic cell death program if concurrent anti-apoptotic NF-B activation is absent (4).
The TGF␤-activated kinase 1 (TAK1) complex has a central function in many inflammatory pathways such as the TNFR, interleukin 1␤ receptor (IL-1R), and several Toll-like receptors (TLRs) (5,6). This complex contains the catalytic subunit TAK1 and the adaptors TAK1-binding proteins 1, 2, and 3 (TAB1, TAB2, TAB3). Upon TNF stimulation, this complex interacts with TRAF2 leading to ubiquitin-mediated activation of TAK1. TAK1, in turn, phosphorylates and activates IB kinase ␣ and ␤ (IKK␣ and IKK␤) and the MAPK kinases (MKK) (7). IKK phosphorylates the NF-B inhibitor IB leading to its degradation and to the subsequent release and nuclear translocation of NF-B. In addition, phosphorylation of MKK4/7 and MKK3/6 by TAK1 leads to the activation of JNK (c-Jun N-terminal kinase) and of the p38 MAPKs. Activation of NF-B and MAPK then coordinately promotes the expression of a wide variety of genes such as IL-6 and IL-8 that contribute to the proinflammatory role of TNF (2,8).
Although the understanding of the TNF signaling cascade has advanced, many of the molecular mechanisms that regulate this pathway remain unclear. Here, we report the identification and characterization of a novel regulator of the TNF pathway, transmembrane protein 9B (TMEM9B, c11orf15). This molecule has been discovered previously as an NF-B inducer in large scale cDNA overexpression screens (9). 3 The function of TMEM9B and of its closest homolog, TMEM9 (10), is not known. Our results indicate that TMEM9B is a lysosomal transmembrane protein that regulates cytokine production induced * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1
Cell Culture and Transfection-The cell lines 293-EBNA, ME180 (both from Invitrogen), and HeLa (from ATCC) were maintained in DMEM (Invitrogen) supplemented with 10% fetal calf serum. Normal human dermal fibroblasts from neonatal skin (Lonza) were cultured in FGM-2 (Lonza) for less than 10 passages. Unless indicated otherwise, cells were stimulated with TNF at 30 ng/ml, IL-1␤ at 10 ng/ml, LPS at 10 ng/ml, poly(I:C) at 0.5 g/ml, and Pam3Cys at 1 g/ml. For siRNA transfection, HeLa, ME180, or primary human dermal fibroblasts were seeded in 150 l of culture medium in 96-well plates. Then 50 l of Opti-MEM (Invitrogen) containing 20 nM siRNA and 0.75 l of HiPerFect (Qiagen) were added to the cells. For transfection of 293-EBNA with DNA, cells were seeded in 100 l of culture medium, and then 10 l of Opti-MEM containing 0.8 l of FuGENE (Roche Applied Science) and 75 ng of luciferase reporter plasmid, 5 ng of pRL-TK, and 120 ng of expression plasmid were added to the cells. For sequential transfection of HeLa cells with siRNA and DNA, cells were seeded in 2 ml of culture medium in 6-well plates and then 100 l of Opti-MEM containing 120 nM siRNA and 6 l of HiPerFect were added to the cells. Two days later, cells were trypsinized and replated in a 96-well plate. Four hours later, the culture medium was replaced by 100 l of Opti-MEM containing 100 ng of plasmid DNA and 1 l of GenePORTER (Genlantis). Transfection medium was replaced three hours later by DMEM, 2% fetal calf serum.
Immunoblotting, Pulldown, and Subcellular Fractionation-For analysis of phosphoproteins, cells were stimulated with TNF for the indicated amount of time, washed once with icecold PBS, and lysed in 50 mM Tris, pH 7.4, 5 mM EDTA, 1% SDS. B, the hydrophilic and membrane protein fractions of HeLa cells were isolated by phase partitioning. The fractions were adjusted to equal volume and subjected to immunoblot. C, biotinylation of cell surface proteins of HeLa cells was performed by incubation with hydrophilic reactive biotin at 4°C. After lysis, biotinylated proteins were isolated on streptavidin-agarose beads and analyzed by immunoblot with anti-TMEM9B or anti-TNFR1 antibodies.
Lysates were heated at 95°C for 5 min and sonicated, and protein concentration was determined with a bicinchoninic acid assay (Sigma). Equal amounts of proteins were resolved on 4 -12% polyacrylamide NuPAGE gels (Invitrogen) and analyzed by immunoblotting. For isolation of the TNFR1 complex, HeLa cells were incubated with cold DMEM containing 100 ng/ml biotinylated TNF for 10 min. Excess biotinyl-TNF was washed off with ice-cold PBS, and then the cells were incubated with prewarmed DMEM at 37°C for the indicated time and rinsed with ice-cold PBS. Cells were lysed in lysis buffer A containing 20 mM Tris pH 7.4, 150 mM NaCl, 1% Triton X-100, 10 mM ␤-glycerol-phosphate, 1 mM EDTA, 1 mM EGTA, 10 mM NaF, and protease inhibitors (Roche Applied Science). After being precleared with protein G-agarose (Roche Applied Science), lysates were incubated with streptavidin-agarose beads (Sigma) for 3 h. Beads were washed five times with lysis buffer, and bound proteins were analyzed by immunoblot as described above. Nuclei were isolated using NE-PER reagent, and integral membrane proteins were isolated by phase partitioning using the Mem-PER reagent as recommended by the manufacturer (Pierce).
Cell Surface Biotinylation-HeLa cells were incubated with 0.2 mg/ml NHS-PEO4-Biotin (Pierce) in PBS for 30 min at 4°C under gentle rocking. Cells were washed with cold PBS, incubated for 30 min at 4°C with DMEM to quench unreacted biotin, and washed again with cold PBS. Cells were then lysed in lysis buffer A and incubated with streptavidin-agarose beads for 2 h. Bound proteins were analyzed by immunoblot as described above.
IL-6, IL-8, Cell Viability, and Reporter Gene Assay-The concentrations of IL-6 or IL-8 were determined in 10 l of cell culture supernatant by HTRF assays as recommended by the manufacturer (Cisbio). Cell viability was assessed with ATPlite 1Step (PerkinElmer Life Sciences). Promoter activity of IL-8 and NF-B reporter plasmids in 293-EBNA cells was assayed with a dual-luciferase reporter assay system (Promega). The ratio of firefly luciferase activity to Renilla luciferase activity was used to normalize for transfection variability.
Quantitative PCR-For the determination of IL-6 and IL-8 mRNA levels, total RNA was isolated on RNeasy micro-columns (Qiagen), and cDNA was synthesized using Superscript III (Invitrogen). On aliquots of this cDNA, quantitative real-time PCR was performed simultaneously with primers and probes (TaqMan Assays-on-Demand) for S18 ribosomal RNA (VIC-labeled) and for IL-6 or IL-8 (both FAM-labeled) with the Taq-Man Fast Universal PCR mix (all from Applied Biosystems). IL-6 and IL-8 mRNA levels were normalized to S18 rRNA levels. Confocal Microscopy-HeLa cells stably expressing a TMEM9B-EGFP fusion protein were grown on coverslips and fixed with 4% formaldehyde in PBS. Fixed cells were permeabilized in 0.1% saponin, 3% horse serum in PBS and then stained sequentially with primary and secondary antibodies and with 4,6-diamidino-2-phenylindole (1 g/ml). Images were obtained and processed using an LSM510 fluorescent confocal microscope (Carl Zeiss).
Statistical Analysis-Statistical significance was determined by Student's t test for two-group comparison and by analysis of variance for multiple group comparison using InStat (Graph-Pad Software).

TMEM9B Is a Glycosylated Lysosomal Transmembrane
Protein-The amino acid sequence of TMEM9B has been shown to contain an N-terminal signal peptide at positions 1-33 (11). Cleavage of the signal peptide gives rise to mature TMEM9B (molecular mass, 19.0 kDa), which is predicted to be glycosylated at Asn-60. Incubation of HeLa cells with tunicamycin, an inhibitor of N-glycosylation, led to a decrease of the apparent molecular mass of TMEM9B from about 22 to 19 kDa (Fig. 1A), confirming that TMEM9B is an N-glycosylated protein.
Sequence analysis indicated that TMEM9B may contain a transmembrane helix located in the signal peptide at position 7-29 and a second transmembrane domain at position 103-125 (TMHMM server v. 2.0 (12). To determine whether TMEM9B is a membrane spanning protein, HeLa cell proteins were fractionated by phase partitioning and analyzed by immunoblot. A strong enrichment of TMEM9B in the hydrophobic fraction mainly containing membrane proteins could be detected, suggesting that TMEM9B is a membrane protein (Fig.  1B). To determine whether TMEM9B is localized at the plasma membrane and is accessible at the outer membrane leaflet, cell surface proteins were biotinylated. HeLa cells were incubated with amine-reactive hydrophilic biotin, and biotinylated proteins were isolated on streptavidin beads. In contrast to the positive control TNFR1, TMEM9B could not be detected in the bound fraction (Fig. 1C), indicating that TMEM9B is not accessible at the outer plasma membrane but that it is an intracellular membrane protein. To assess whether TMEM9B is a secreted protein, tagged TMEM9B was transiently overexpressed in HeLa cells, and after 24 h supernatant was analyzed by immunoprecipitation for the presence of soluble tagged TMEM9B. No secreted TMEM9B could be detected (data not shown).
To investigate its subcellular localization, a stable HeLa cell clone expressing TMEM9B fused to EGFP was analyzed. No TMEM9B-EGFP could be detected at the plasma membrane of live cells by confocal microscopy ( Fig. 2A), confirming the result of the cell surface biotinylation assay. A distinct perinuclear and vesicular structure was however revealed by the TMEM9B-EGFP fluorescence ( Fig. 2A). To determine the nature of this structure, double staining with markers for lysosomes, early endosomes, and the Golgi apparatus was performed in fixed cells. Fig. 2B shows strong co-localization (in yellow) of TMEM9B with the lysosomal marker LAMP1 and partial colocalization with the early endosomal marker, EEA1. In contrast, no colocalization with the Golgi protein mannosidase could be observed. Taken together, these data show that TMEM9B is a transmembrane protein localized primarily in the lysosome and, to a limited extent, in early endosomes.
TMEM9B Is Required for Cytokine Production in Response to Several Inflammatory Stimuli-The two proinflammatory cytokines IL-6 and IL-8 are induced by TNF in many different cell types. To determine whether TMEM9B is involved in the TNF signaling cascade, we first assessed whether the expression of TMEM9B was necessary for the production of IL-6 and IL-8 in HeLa cells. Endogenous TMEM9B was depleted with specific siRNA, and IL-6 and IL-8 were measured in the cell culture supernatant. Two different siRNAs against TMEM9B reduced the production of IL-6 induced by TNF stimulation by 61 and 35% and of IL-8 by 47 and 24% (Fig. 3A). The effect of the siRNAs on cytokine secretion correlated with an efficient reduction of TMEM9B protein expression (Fig. 3B).
The role of TMEM9B in the TNF signaling cascade was further confirmed in primary human dermal fibroblasts. Similarly to HeLa cells, an siRNA against TMEM9B inhibited the production of IL-8 induced by TNF by 62% (Fig. 4A) with a concomitant reduction of TMEM9B protein levels (Fig. 4B). We next determined whether TMEM9B was involved in other inflammatory signaling cascades in primary human dermal fibroblasts. Depletion of TMEM9B reduced IL-1␤-induced IL-8 by 56% and the IL-8 production induced by ligands of TLR2 (Pam3Cys), TLR3 (poly(I:C)), and TLR4 (LPS) by 58, 68, and 77%, respectively (Fig. 4A). These data show that TMEM9B is required for the production of proinflammatory cytokines not only downstream of the TNFR1 but also downstream of the IL-1 receptor and several TLRs, suggesting that TMEM9B regulates a common proinflammatory signaling module.

TMEM9B Is Not Required for TNF or Fas Ligand-induced
Apoptosis-In addition to promote the expression of proinflammatory cytokines, TNF also triggers apoptotic cell death. We sought therefore to determine whether TMEM9B regulated TNF-induced apoptosis. The ME180 cell line is sensitive to TNF-induced apoptosis (13), and treatment with TNF resulted in about 20 and 30% dead cells (after 10 and 100 ng/ml TNF, respectively; Fig. 5A). Cell death was TNF-specific, as it was blocked by an siRNA against TNFR1. In contrast, reduction of TMEM9B expression in ME180 cells with an siRNA (Fig. 5B) did not significantly inhibit cell death compared with cells transfected with the control siRNA.  The HeLa cell line is not sensitive to TNF-induced apoptosis, but it is sensitive to Fas ligand, a related member of the TNF family. Stimulation with 200 ng/ml hexameric super-Fas ligand killed about 50% of the cells after 16 h (Fig. 5C). Knockdown of TMEM9B did not affect Fas ligand-induced cell death. Together, these data demonstrate that the induction of apoptosis by the TNF family members TNF and Fas ligand is independent of TMEM9B and that TMEM9B is specifically involved in inflammatory cytokine signaling.
TMEM9B Regulates Cytokine Expression at the Transcriptional Level-The reduction in cytokine production after TMEM9B depletion could result from a defect in gene transcription, mRNA translation, or protein secretion. We there-fore assessed whether cytokine mRNA levels were affected by TMEM9B. The IL-6 and IL-8 mRNA levels of HeLa cells transfected with TMEM9B siRNA were determined by quantitative PCR. The mRNA expression of both cytokines was rapidly induced upon TNF stimulation, and the maximum was reached after 60 min (Fig. 6A). An siRNA against TMEM9B almost completely abrogated both IL-6 and IL-8 induction. The role of  TMEM9B on IL-8 promoter activity was determined by cotransfection of a TMEM9B expression vector and an IL-8 promoter reporter. Overexpression of TMEM9B activated the IL-8 promoter in a dose-dependent manner to a level similar to that observed upon TNF stimulation (Fig. 6B). Together, these data show that TMEM9B is both necessary and sufficient for TNF-induced cytokine gene transcription.
TMEM9B Acts Downstream of the TNFR1 Complex and of RIP1-Upon ligand binding, TNFR1 recruits TRADD and RIP1 to form a complex that is internalized into signaling vesicles termed TNF receptosomes (14). To determine whether TMEM9B is recruited to TNF receptosomes, we isolated the active TNFR1 complex using biotinylated TNF. Incubation of HeLa cells with biotinyl-TNF lead to the recruitment of TRADD as well as of polyubiquitinated RIP1 (Fig. 7A). Because TMEM9B could not be detected in the biotinyl-TNF pulldown, it likely is not recruited to the TNFR1 complex but probably acts further downstream of the TNFR1 complex.
Polyubiquitinated RIP1 has been shown to serve as a platform to recruit the TAK1 and IKK complexes, leading to the activation of the IKK by TAK1 (15). We sought to determine whether TMEM9B acts upstream or downstream of RIP1 and TAK1. Therefore, HeLa cells were sequentially transfected, first with siRNAs against TMEM9B or p65 and then 2 days later with expression plasmids for RIP1 or TAK1 and TAB1. Overexpression of RIP1 or co-expression of TAK1 with TAB1 induced IL-8 production (Fig. 7B). In cells transfected with an siRNA against TMEM9B or p65, IL-8 production was strongly reduced whether it was induced by RIP1 or TAK1/TAB1 overexpression. These data strongly suggest that TMEM9B acts downstream of RIP1 and downstream or at the level of TAK1.
TMEM9B Is Required for TNF-induced NF-B Signaling-Overexpression of TMEM9B has been reported to activate an NF-B promoter reporter vector (9). In a manner similar to our results with the IL-8 promoter reporter (Fig. 6B), we confirmed the activity on an NF-B promoter in 293-EBNA cells cotransfected with a TMEM9B expression vector and an NF-B reporter plasmid (Fig. 8A). We next determined at which level TMEM9B was involved in the NF-B signaling cascade, first assessing whether TMEM9B was required for the nuclear translocation of p65. TNF stimulation induced nuclear translocation of p65 after 30 min in control cells (Fig.  8B). In cells transfected with a siRNA against TMEM9B, the levels of nuclear p65 were reduced by about 50% compared with controls. As phosphorylation of p65 is necessary for its full transcriptional activation (16), we next determined whether depletion of TMEM9B affected the phosphorylation of p65 upon TNF stimulation. The phosphorylation of p65 could be detected within 3 min after TNF stimulation and was sustained for at least 10 min in control cells (Fig. 8C). In cells transfected with an siRNA against TMEM9B, p65 phosphorylation was moderately reduced at 3 min and completely blocked at 10 min. These data suggest that TMEM9B is required for the translocation of p65 and to sustain its phosphorylation after TNF stimulation, and that TMEM9B acts upstream of p65. We therefore analyzed whether TMEM9B acted upstream or downstream of the IKKs by assessing the phosphorylation state of IKK␣ and IKK␤. Depletion of TMEM9B resulted in a strongly reduced phosphorylation of IKK␣/␤ (Fig. 8C) without affecting the expression level of IKK␣ and IKK␤ (Fig. 8F). This indicates that TMEM9B modulates the NF-B pathway by acting upstream of the IKKs.
TMEM9B Is Required for Activation of the JNK and p38 MAPK Signaling Cascades-In addition to activating the NF-B pathway, TNF also triggers the p38 and JNK signaling cascades via RIP1 and TAK1. In HeLa cells TNF induced the phosphorylation of the two upstream kinases MKK3 and MKK6 and of p38 and MK2, a p38 substrate. An siRNA against TMEM9B almost completely abrogated phosphorylation of MKK3/6, p38, and MK2 ( Fig. 8D) but did not affect the total amount of MKK6 or p38 (Fig. 8F). Similarly, the phosphorylation but not the total amount of MKK4, as well as the phosphorylation of its substrate, and the two JNK isoforms p46 and p54 were strongly reduced (Fig. 8, E and F). Thus, these data show that in addition to its function in the activation of the NF-B pathway, TMEM9B is also required for the activation of the p38 MAPK and JNK signaling cascades, positioning TMEM9B in a proximal and common position of proinflammatory cytokine signaling pathways.

DISCUSSION
TNF activates a complex signaling network that leads to two major outcomes: cell activation and apoptosis (4). In our attempt to better characterize the TNFR1 cascade, we uncovered a function of TMEM9B as an essential component of several inflammatory pathways. We show that TMEM9B is a lysosomal membrane protein involved in the proinflammatory but not in the apoptotic arm of the TNF signaling cascade at the level of TAK1. The expression of TMEM9B was required for the production of IL-6 and IL-8 induced not only by TNF but also immune stimuli like IL-1␤ and several TLR ligands. Depletion of TMEM9B resulted in reduced transcriptional activation of the IL-8 promoter via NF-B and diminished TNF-mediated MAPK signaling.
The epistasis experiments performed in this study suggest that TMEM9B acts at the level of TAK1. Indeed, the knockdown of TMEM9B inhibited the cytokine induction by RIP1 or TAK1, indicating interference downstream of RIP1 and TAK1. Because reduction of TMEM9B expression also resulted in a reduction of the phosphorylation of the TAK1 substrates IKK␣/␤, MKK3/6, and MKK4, TMEM9B very likely acts at the level of TAK1.
Further evidence for the role of TMEM9B in TAK1 regulation is that TMEM9B is involved not only in the TNF signaling cascade but also downstream of the IL-1R and several TLRs, mirroring the central role of TAK1 in the very same pathways (17,18). The molecules required for the activation of TAK1 differ between these receptors. Whereas TNFR1 signals via TRAF2 and RIP1, the IL-1R and TLRs require MYD88, IRAK4, and TRAF6 (5,19). Precisely how TMEM9B controls TAK1 is unclear; TMEM9B may be required either for the phosphorylation and ubiquitin-mediated activation of TAK1 or for the recruitment of TAK1 substrates to the TAK1-TAB complex.
We observed that TMEM9B is localized to vesicular structures of the lysosome. The lack of adequate reagents precluded an analysis of the subcellular localization of endogenous TMEM9B, but a stable low expressing clone of a TMEM9B-EGFP fusion could be used. With this caveat, our data on TMEM9B are reminiscent of the subcellular localization of TMEM9, the closest homolog of TMEM9B, which can be found in lysosomes and late endosomes when overexpressed in COS cells (10). The function of TMEM9, however, is unknown, although a role in intracellular transport was suggested.
The localization of TMEM9B in lysosomes and early endosomes suggests that these organelles are involved in the regulation of signal transduction downstream of inflammatory receptors. Interestingly, the activation of mitogen-activated protein kinase kinase 1 (MEK1) and its downstream substrate, extracellular-signal regulated protein kinase 1/2 (ERK1/2), by epidermal growth factor requires localization of MEK1 and ERK1/2 to the endosome via the endosomal protein p14. Binding of p14 to the scaffold protein MEK partner 1 (MP1) recruits MEK1 (20). It has been shown recently that p14 deficiency abrogates late endosomal biogenesis and ERK1/2 activation, leading to a new type of human primary immunodeficiency (21). Our data showing the involvement of a lysosomal protein in the activation of the NF-B and MAPK pathways suggest by analogy that the lysosomal and/or endosomal compartments may play a central role also in inflammatory signaling. One might speculate that TMEM9B, like p14, could act as an anchor for a signaling complex upstream of NF-B and p38. Interestingly, the TAK1-TAB complex is associated to a membrane structure in unstimulated cells and translocates to the cytosol upon IL-1␤ stimulation (22). Whether the TAK1-TAB complex associates with the lysosomal membrane and whether this is mediated by TMEM9B are currently under investigation. A direct interac-FIGURE 8. TMEM9B is required for NF-B and MAPK activation and acts upstream of IKK and MKKs. A, 293-EBNA cells were transfected with an NF-B promoter luciferase reporter vector and increasing amounts of a TMEM9B expression vector. One day later, luciferase activity was determined. Data are the average of three independent experiments. B, HeLa cells were transfected with a siRNA against TMEM9B or a control siRNA and were stimulated 2 days later with TNF for 30 min. Nuclear proteins were isolated and analyzed for p65 levels. C, HeLa cells were transfected with an siRNA against TMEM9B or a control siRNA and were stimulated 2 days later for the indicated time. Cells were lysed, and phosphorylation of p65 and IKK was assessed by immunoblot. D and E, HeLa cells were transfected with an siRNA against TMEM9B or a control siRNA and were stimulated 2 days later for the indicated time. Cells were lysed, and phosphorylation of MKK3, MKK6, MKK4, p38, MK2, and JNK was assessed by immunoblot. F, HeLa cells were transfected with an siRNA against TMEM9B or a control siRNA, and the expression the expression level of the indicated proteins was assessed by immunoblot. GAPDH, glyceraldehyde-3-phosphate dehydrogenase. tion between TMEM9B and TAK1 or any TAB could not be detected (data not shown), suggesting that TMEM9B may modulate TAK1 indirectly.
Our observations are compatible with the notion that a first membrane-proximal TNFR1 complex exists that signals to NF-B and MAPK prior to an internalized TNFR1 receptosome which mediates apoptotic signaling (14,23). Depletion of TMEM9B did not affect TNF or Fas ligand-induced apoptosis or surface TNFR1 binding to biotinyl-TNF (data not shown). In addition, we did not detect a direct interaction between the TNFR1 complex and TMEM9B. Therefore, our data suggest that TMEM9B is required for the activation of a TAK1 signaling complex that is downstream of the membrane-proximal TNFR1 complex and receptosome and is shared with several inflammatory signals.
In summary, we have described TMEM9B as a glycosylated membrane protein located in the lysosome, demonstrating that TMEM9B is required for the activation of the NF-B and MAPK pathways by inflammatory stimuli. This is the first report to characterize TMEM9B function and also to indicate the relevance of a lysosomal protein in the signaling cascades of inflammatory receptors.